|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 0.19230769230769232, |
|
"eval_steps": 500, |
|
"global_step": 100, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.0, |
|
"learning_rate": 6.25e-07, |
|
"loss": 3.7473, |
|
"step": 1 |
|
}, |
|
{ |
|
"epoch": 0.0, |
|
"learning_rate": 1.25e-06, |
|
"loss": 0.0, |
|
"step": 2 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"learning_rate": 1.8750000000000003e-06, |
|
"loss": 0.0, |
|
"step": 3 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"learning_rate": 2.5e-06, |
|
"loss": 0.0, |
|
"step": 4 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"learning_rate": 3.125e-06, |
|
"loss": 0.0, |
|
"step": 5 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"learning_rate": 3.7500000000000005e-06, |
|
"loss": 0.0, |
|
"step": 6 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"learning_rate": 4.3750000000000005e-06, |
|
"loss": 0.0, |
|
"step": 7 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"learning_rate": 5e-06, |
|
"loss": 0.0, |
|
"step": 8 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"learning_rate": 5.625e-06, |
|
"loss": 0.0, |
|
"step": 9 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"learning_rate": 6.25e-06, |
|
"loss": 0.0, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"learning_rate": 6.875e-06, |
|
"loss": 0.0, |
|
"step": 11 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"learning_rate": 7.500000000000001e-06, |
|
"loss": 0.0, |
|
"step": 12 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"learning_rate": 8.125000000000001e-06, |
|
"loss": 0.0, |
|
"step": 13 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"learning_rate": 8.750000000000001e-06, |
|
"loss": 0.0, |
|
"step": 14 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"learning_rate": 9.375000000000001e-06, |
|
"loss": 0.0, |
|
"step": 15 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"learning_rate": 1e-05, |
|
"loss": 0.0, |
|
"step": 16 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"learning_rate": 1.0625e-05, |
|
"loss": 0.0, |
|
"step": 17 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"learning_rate": 1.125e-05, |
|
"loss": 0.0, |
|
"step": 18 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"learning_rate": 1.1875e-05, |
|
"loss": 0.0, |
|
"step": 19 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"learning_rate": 1.25e-05, |
|
"loss": 0.0, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"learning_rate": 1.3125e-05, |
|
"loss": 0.0, |
|
"step": 21 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"learning_rate": 1.375e-05, |
|
"loss": 0.0, |
|
"step": 22 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"learning_rate": 1.4375e-05, |
|
"loss": 0.0, |
|
"step": 23 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"learning_rate": 1.5000000000000002e-05, |
|
"loss": 0.0, |
|
"step": 24 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"learning_rate": 1.5625e-05, |
|
"loss": 0.0, |
|
"step": 25 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"learning_rate": 1.6250000000000002e-05, |
|
"loss": 0.0, |
|
"step": 26 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"learning_rate": 1.6875e-05, |
|
"loss": 0.0, |
|
"step": 27 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"learning_rate": 1.7500000000000002e-05, |
|
"loss": 0.0, |
|
"step": 28 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"learning_rate": 1.8125e-05, |
|
"loss": 0.0, |
|
"step": 29 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"learning_rate": 1.8750000000000002e-05, |
|
"loss": 0.0, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"learning_rate": 1.9375e-05, |
|
"loss": 0.0, |
|
"step": 31 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 32 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"learning_rate": 1.9999951432210905e-05, |
|
"loss": 0.0, |
|
"step": 33 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"learning_rate": 1.9999805729315383e-05, |
|
"loss": 0.0, |
|
"step": 34 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"learning_rate": 1.999956289272873e-05, |
|
"loss": 0.0, |
|
"step": 35 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"learning_rate": 1.999922292480975e-05, |
|
"loss": 0.0, |
|
"step": 36 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"learning_rate": 1.9998785828860744e-05, |
|
"loss": 0.0, |
|
"step": 37 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"learning_rate": 1.9998251609127465e-05, |
|
"loss": 0.0, |
|
"step": 38 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"learning_rate": 1.999762027079909e-05, |
|
"loss": 0.0, |
|
"step": 39 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"learning_rate": 1.9996891820008165e-05, |
|
"loss": 0.0, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"learning_rate": 1.9996066263830533e-05, |
|
"loss": 0.0, |
|
"step": 41 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"learning_rate": 1.9995143610285275e-05, |
|
"loss": 0.0, |
|
"step": 42 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"learning_rate": 1.9994123868334655e-05, |
|
"loss": 0.0, |
|
"step": 43 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"learning_rate": 1.9993007047883988e-05, |
|
"loss": 0.0, |
|
"step": 44 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"learning_rate": 1.999179315978157e-05, |
|
"loss": 0.0, |
|
"step": 45 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"learning_rate": 1.999048221581858e-05, |
|
"loss": 0.0, |
|
"step": 46 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"learning_rate": 1.9989074228728942e-05, |
|
"loss": 0.0, |
|
"step": 47 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"learning_rate": 1.9987569212189224e-05, |
|
"loss": 0.0, |
|
"step": 48 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"learning_rate": 1.9985967180818493e-05, |
|
"loss": 0.0, |
|
"step": 49 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"learning_rate": 1.998426815017817e-05, |
|
"loss": 0.0, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"learning_rate": 1.998247213677188e-05, |
|
"loss": 0.0, |
|
"step": 51 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"learning_rate": 1.9980579158045322e-05, |
|
"loss": 0.0, |
|
"step": 52 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"learning_rate": 1.9978589232386036e-05, |
|
"loss": 0.0, |
|
"step": 53 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"learning_rate": 1.997650237912329e-05, |
|
"loss": 0.0, |
|
"step": 54 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"learning_rate": 1.997431861852785e-05, |
|
"loss": 0.0, |
|
"step": 55 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"learning_rate": 1.9972037971811802e-05, |
|
"loss": 0.0, |
|
"step": 56 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"learning_rate": 1.996966046112834e-05, |
|
"loss": 0.0, |
|
"step": 57 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"learning_rate": 1.996718610957155e-05, |
|
"loss": 0.0, |
|
"step": 58 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"learning_rate": 1.9964614941176194e-05, |
|
"loss": 0.0, |
|
"step": 59 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"learning_rate": 1.9961946980917457e-05, |
|
"loss": 0.0, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"learning_rate": 1.995918225471073e-05, |
|
"loss": 0.0, |
|
"step": 61 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"learning_rate": 1.9956320789411338e-05, |
|
"loss": 0.0, |
|
"step": 62 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"learning_rate": 1.9953362612814294e-05, |
|
"loss": 0.0, |
|
"step": 63 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"learning_rate": 1.9950307753654016e-05, |
|
"loss": 0.0, |
|
"step": 64 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"learning_rate": 1.994715624160405e-05, |
|
"loss": 0.0, |
|
"step": 65 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"learning_rate": 1.99439081072768e-05, |
|
"loss": 0.0, |
|
"step": 66 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"learning_rate": 1.9940563382223196e-05, |
|
"loss": 0.0, |
|
"step": 67 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"learning_rate": 1.9937122098932428e-05, |
|
"loss": 0.0, |
|
"step": 68 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"learning_rate": 1.9933584290831593e-05, |
|
"loss": 0.0, |
|
"step": 69 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"learning_rate": 1.9929949992285397e-05, |
|
"loss": 0.0, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"learning_rate": 1.992621923859581e-05, |
|
"loss": 0.0, |
|
"step": 71 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"learning_rate": 1.9922392066001724e-05, |
|
"loss": 0.0, |
|
"step": 72 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"learning_rate": 1.99184685116786e-05, |
|
"loss": 0.0, |
|
"step": 73 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"learning_rate": 1.9914448613738107e-05, |
|
"loss": 0.0, |
|
"step": 74 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"learning_rate": 1.991033241122776e-05, |
|
"loss": 0.0, |
|
"step": 75 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"learning_rate": 1.9906119944130527e-05, |
|
"loss": 0.0, |
|
"step": 76 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"learning_rate": 1.9901811253364458e-05, |
|
"loss": 0.0, |
|
"step": 77 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"learning_rate": 1.9897406380782262e-05, |
|
"loss": 0.0, |
|
"step": 78 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"learning_rate": 1.989290536917093e-05, |
|
"loss": 0.0, |
|
"step": 79 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"learning_rate": 1.9888308262251286e-05, |
|
"loss": 0.0, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"learning_rate": 1.988361510467761e-05, |
|
"loss": 0.0, |
|
"step": 81 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"learning_rate": 1.9878825942037147e-05, |
|
"loss": 0.0, |
|
"step": 82 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"learning_rate": 1.9873940820849714e-05, |
|
"loss": 0.0, |
|
"step": 83 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"learning_rate": 1.9868959788567213e-05, |
|
"loss": 0.0, |
|
"step": 84 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"learning_rate": 1.9863882893573188e-05, |
|
"loss": 0.0, |
|
"step": 85 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"learning_rate": 1.985871018518236e-05, |
|
"loss": 0.0, |
|
"step": 86 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"learning_rate": 1.9853441713640123e-05, |
|
"loss": 0.0, |
|
"step": 87 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"learning_rate": 1.9848077530122083e-05, |
|
"loss": 0.0, |
|
"step": 88 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"learning_rate": 1.9842617686733546e-05, |
|
"loss": 0.0, |
|
"step": 89 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"learning_rate": 1.9837062236509013e-05, |
|
"loss": 0.0, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"learning_rate": 1.983141123341168e-05, |
|
"loss": 0.0, |
|
"step": 91 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"learning_rate": 1.9825664732332886e-05, |
|
"loss": 0.0, |
|
"step": 92 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"learning_rate": 1.9819822789091597e-05, |
|
"loss": 0.0, |
|
"step": 93 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"learning_rate": 1.981388546043388e-05, |
|
"loss": 0.0, |
|
"step": 94 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"learning_rate": 1.9807852804032306e-05, |
|
"loss": 0.0, |
|
"step": 95 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"learning_rate": 1.9801724878485438e-05, |
|
"loss": 0.0, |
|
"step": 96 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"learning_rate": 1.979550174331724e-05, |
|
"loss": 0.0, |
|
"step": 97 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"learning_rate": 1.9789183458976485e-05, |
|
"loss": 0.0, |
|
"step": 98 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"learning_rate": 1.97827700868362e-05, |
|
"loss": 0.0, |
|
"step": 99 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"learning_rate": 1.977626168919305e-05, |
|
"loss": 0.0, |
|
"step": 100 |
|
} |
|
], |
|
"logging_steps": 1.0, |
|
"max_steps": 1040, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 2, |
|
"save_steps": 100, |
|
"total_flos": 4.989387064056218e+16, |
|
"train_batch_size": 16, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|